| OLD | NEW |
| 1 /* | 1 /* |
| 2 * (C) 1999 Lars Knoll (knoll@kde.org) | 2 * (C) 1999 Lars Knoll (knoll@kde.org) |
| 3 * Copyright (C) 2004, 2005, 2006, 2007, 2008, 2010, 2012 Apple Inc. All rights
reserved. | 3 * Copyright (C) 2004, 2005, 2006, 2007, 2008, 2010, 2012 Apple Inc. All rights
reserved. |
| 4 * Copyright (C) 2007-2009 Torch Mobile, Inc. | 4 * Copyright (C) 2007-2009 Torch Mobile, Inc. |
| 5 * | 5 * |
| 6 * This library is free software; you can redistribute it and/or | 6 * This library is free software; you can redistribute it and/or |
| 7 * modify it under the terms of the GNU Library General Public | 7 * modify it under the terms of the GNU Library General Public |
| 8 * License as published by the Free Software Foundation; either | 8 * License as published by the Free Software Foundation; either |
| 9 * version 2 of the License, or (at your option) any later version. | 9 * version 2 of the License, or (at your option) any later version. |
| 10 * | 10 * |
| (...skipping 24 matching lines...) Expand all Loading... |
| 35 #include "wtf/text/UTF8.h" | 35 #include "wtf/text/UTF8.h" |
| 36 #include "wtf/text/Unicode.h" | 36 #include "wtf/text/Unicode.h" |
| 37 #include <stdarg.h> | 37 #include <stdarg.h> |
| 38 | 38 |
| 39 namespace WTF { | 39 namespace WTF { |
| 40 | 40 |
| 41 using namespace Unicode; | 41 using namespace Unicode; |
| 42 | 42 |
| 43 // Construct a string with UTF-16 data. | 43 // Construct a string with UTF-16 data. |
| 44 String::String(const UChar* characters, unsigned length) | 44 String::String(const UChar* characters, unsigned length) |
| 45 : m_impl(characters ? StringImpl::create(characters, length) : nullptr) | 45 : m_impl(characters ? StringImpl::create(characters, length) : nullptr) { |
| 46 { | |
| 47 } | 46 } |
| 48 | 47 |
| 49 // Construct a string with UTF-16 data, from a null-terminated source. | 48 // Construct a string with UTF-16 data, from a null-terminated source. |
| 50 String::String(const UChar* str) | 49 String::String(const UChar* str) { |
| 51 { | 50 if (!str) |
| 52 if (!str) | 51 return; |
| 53 return; | 52 m_impl = StringImpl::create(str, lengthOfNullTerminatedString(str)); |
| 54 m_impl = StringImpl::create(str, lengthOfNullTerminatedString(str)); | |
| 55 } | 53 } |
| 56 | 54 |
| 57 // Construct a string with latin1 data. | 55 // Construct a string with latin1 data. |
| 58 String::String(const LChar* characters, unsigned length) | 56 String::String(const LChar* characters, unsigned length) |
| 59 : m_impl(characters ? StringImpl::create(characters, length) : nullptr) | 57 : m_impl(characters ? StringImpl::create(characters, length) : nullptr) { |
| 60 { | |
| 61 } | 58 } |
| 62 | 59 |
| 63 String::String(const char* characters, unsigned length) | 60 String::String(const char* characters, unsigned length) |
| 64 : m_impl(characters ? StringImpl::create(reinterpret_cast<const LChar*>(char
acters), length) : nullptr) | 61 : m_impl(characters ? StringImpl::create(reinterpret_cast<const LChar*>(char
acters), length) : nullptr) { |
| 65 { | |
| 66 } | 62 } |
| 67 | 63 |
| 68 // Construct a string with latin1 data, from a null-terminated source. | 64 // Construct a string with latin1 data, from a null-terminated source. |
| 69 String::String(const LChar* characters) | 65 String::String(const LChar* characters) |
| 70 : m_impl(characters ? StringImpl::create(characters) : nullptr) | 66 : m_impl(characters ? StringImpl::create(characters) : nullptr) { |
| 71 { | |
| 72 } | 67 } |
| 73 | 68 |
| 74 String::String(const char* characters) | 69 String::String(const char* characters) |
| 75 : m_impl(characters ? StringImpl::create(reinterpret_cast<const LChar*>(char
acters)) : nullptr) | 70 : m_impl(characters ? StringImpl::create(reinterpret_cast<const LChar*>(char
acters)) : nullptr) { |
| 76 { | 71 } |
| 77 } | 72 |
| 78 | 73 void String::append(const String& string) { |
| 79 void String::append(const String& string) | 74 if (string.isEmpty()) |
| 80 { | 75 return; |
| 81 if (string.isEmpty()) | 76 if (!m_impl) { |
| 82 return; | 77 m_impl = string.m_impl; |
| 83 if (!m_impl) { | 78 return; |
| 84 m_impl = string.m_impl; | 79 } |
| 85 return; | 80 |
| 86 } | 81 // FIXME: This is extremely inefficient. So much so that we might want to |
| 87 | 82 // take this out of String's API. We can make it better by optimizing the |
| 88 // FIXME: This is extremely inefficient. So much so that we might want to | 83 // case where exactly one String is pointing at this StringImpl, but even |
| 89 // take this out of String's API. We can make it better by optimizing the | 84 // then it's going to require a call into the allocator every single time. |
| 90 // case where exactly one String is pointing at this StringImpl, but even | 85 |
| 91 // then it's going to require a call into the allocator every single time. | 86 if (m_impl->is8Bit() && string.m_impl->is8Bit()) { |
| 92 | 87 LChar* data; |
| 93 if (m_impl->is8Bit() && string.m_impl->is8Bit()) { | |
| 94 LChar* data; | |
| 95 RELEASE_ASSERT(string.length() <= std::numeric_limits<unsigned>::max() -
m_impl->length()); | |
| 96 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(m_impl->len
gth() + string.length(), data); | |
| 97 memcpy(data, m_impl->characters8(), m_impl->length() * sizeof(LChar)); | |
| 98 memcpy(data + m_impl->length(), string.characters8(), string.length() *
sizeof(LChar)); | |
| 99 m_impl = newImpl.release(); | |
| 100 return; | |
| 101 } | |
| 102 | |
| 103 UChar* data; | |
| 104 RELEASE_ASSERT(string.length() <= std::numeric_limits<unsigned>::max() - m_i
mpl->length()); | 88 RELEASE_ASSERT(string.length() <= std::numeric_limits<unsigned>::max() - m_i
mpl->length()); |
| 105 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(m_impl->length(
) + string.length(), data); | 89 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(m_impl->length(
) + string.length(), data); |
| 106 | 90 memcpy(data, m_impl->characters8(), m_impl->length() * sizeof(LChar)); |
| 107 if (m_impl->is8Bit()) | 91 memcpy(data + m_impl->length(), string.characters8(), string.length() * size
of(LChar)); |
| 108 StringImpl::copyChars(data, m_impl->characters8(), m_impl->length()); | |
| 109 else | |
| 110 StringImpl::copyChars(data, m_impl->characters16(), m_impl->length()); | |
| 111 | |
| 112 if (string.impl()->is8Bit()) | |
| 113 StringImpl::copyChars(data + m_impl->length(), string.impl()->characters
8(), string.impl()->length()); | |
| 114 else | |
| 115 StringImpl::copyChars(data + m_impl->length(), string.impl()->characters
16(), string.impl()->length()); | |
| 116 | |
| 117 m_impl = newImpl.release(); | 92 m_impl = newImpl.release(); |
| 93 return; |
| 94 } |
| 95 |
| 96 UChar* data; |
| 97 RELEASE_ASSERT(string.length() <= std::numeric_limits<unsigned>::max() - m_imp
l->length()); |
| 98 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(m_impl->length()
+ string.length(), data); |
| 99 |
| 100 if (m_impl->is8Bit()) |
| 101 StringImpl::copyChars(data, m_impl->characters8(), m_impl->length()); |
| 102 else |
| 103 StringImpl::copyChars(data, m_impl->characters16(), m_impl->length()); |
| 104 |
| 105 if (string.impl()->is8Bit()) |
| 106 StringImpl::copyChars(data + m_impl->length(), string.impl()->characters8(),
string.impl()->length()); |
| 107 else |
| 108 StringImpl::copyChars(data + m_impl->length(), string.impl()->characters16()
, string.impl()->length()); |
| 109 |
| 110 m_impl = newImpl.release(); |
| 118 } | 111 } |
| 119 | 112 |
| 120 template <typename CharacterType> | 113 template <typename CharacterType> |
| 121 inline void String::appendInternal(CharacterType c) | 114 inline void String::appendInternal(CharacterType c) { |
| 122 { | 115 // FIXME: This is extremely inefficient. So much so that we might want to |
| 123 // FIXME: This is extremely inefficient. So much so that we might want to | 116 // take this out of String's API. We can make it better by optimizing the |
| 124 // take this out of String's API. We can make it better by optimizing the | 117 // case where exactly one String is pointing at this StringImpl, but even |
| 125 // case where exactly one String is pointing at this StringImpl, but even | 118 // then it's going to require a call into the allocator every single time. |
| 126 // then it's going to require a call into the allocator every single time. | 119 if (!m_impl) { |
| 127 if (!m_impl) { | 120 m_impl = StringImpl::create(&c, 1); |
| 128 m_impl = StringImpl::create(&c, 1); | 121 return; |
| 129 return; | 122 } |
| 130 } | 123 |
| 131 | 124 // FIXME: We should be able to create an 8 bit string via this code path. |
| 132 // FIXME: We should be able to create an 8 bit string via this code path. | 125 UChar* data; |
| 133 UChar* data; | 126 RELEASE_ASSERT(m_impl->length() < std::numeric_limits<unsigned>::max()); |
| 134 RELEASE_ASSERT(m_impl->length() < std::numeric_limits<unsigned>::max()); | 127 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(m_impl->length()
+ 1, data); |
| 135 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(m_impl->length(
) + 1, data); | 128 if (m_impl->is8Bit()) |
| 136 if (m_impl->is8Bit()) | 129 StringImpl::copyChars(data, m_impl->characters8(), m_impl->length()); |
| 137 StringImpl::copyChars(data, m_impl->characters8(), m_impl->length()); | 130 else |
| 138 else | 131 StringImpl::copyChars(data, m_impl->characters16(), m_impl->length()); |
| 139 StringImpl::copyChars(data, m_impl->characters16(), m_impl->length()); | 132 data[m_impl->length()] = c; |
| 140 data[m_impl->length()] = c; | 133 m_impl = newImpl.release(); |
| 141 m_impl = newImpl.release(); | 134 } |
| 142 } | 135 |
| 143 | 136 void String::append(LChar c) { |
| 144 void String::append(LChar c) | 137 appendInternal(c); |
| 145 { | 138 } |
| 146 appendInternal(c); | 139 |
| 147 } | 140 void String::append(UChar c) { |
| 148 | 141 appendInternal(c); |
| 149 void String::append(UChar c) | 142 } |
| 150 { | 143 |
| 151 appendInternal(c); | 144 int codePointCompare(const String& a, const String& b) { |
| 152 } | 145 return codePointCompare(a.impl(), b.impl()); |
| 153 | 146 } |
| 154 int codePointCompare(const String& a, const String& b) | 147 |
| 155 { | 148 void String::insert(const String& string, unsigned position) { |
| 156 return codePointCompare(a.impl(), b.impl()); | 149 if (string.isEmpty()) { |
| 157 } | 150 if (string.isNull()) |
| 158 | 151 return; |
| 159 void String::insert(const String& string, unsigned position) | 152 if (isNull()) |
| 160 { | 153 m_impl = string.impl(); |
| 161 if (string.isEmpty()) { | 154 return; |
| 162 if (string.isNull()) | 155 } |
| 163 return; | 156 |
| 164 if (isNull()) | 157 if (string.is8Bit()) |
| 165 m_impl = string.impl(); | 158 insert(string.impl()->characters8(), string.length(), position); |
| 166 return; | 159 else |
| 167 } | 160 insert(string.impl()->characters16(), string.length(), position); |
| 168 | 161 } |
| 169 if (string.is8Bit()) | 162 |
| 170 insert(string.impl()->characters8(), string.length(), position); | 163 void String::append(const LChar* charactersToAppend, unsigned lengthToAppend) { |
| 171 else | 164 if (!m_impl) { |
| 172 insert(string.impl()->characters16(), string.length(), position); | 165 if (!charactersToAppend) |
| 173 } | 166 return; |
| 174 | 167 m_impl = StringImpl::create(charactersToAppend, lengthToAppend); |
| 175 void String::append(const LChar* charactersToAppend, unsigned lengthToAppend) | 168 return; |
| 176 { | 169 } |
| 177 if (!m_impl) { | 170 |
| 178 if (!charactersToAppend) | 171 if (!lengthToAppend) |
| 179 return; | 172 return; |
| 180 m_impl = StringImpl::create(charactersToAppend, lengthToAppend); | 173 |
| 181 return; | 174 ASSERT(charactersToAppend); |
| 182 } | 175 |
| 183 | 176 unsigned strLength = m_impl->length(); |
| 184 if (!lengthToAppend) | 177 |
| 185 return; | 178 if (m_impl->is8Bit()) { |
| 186 | |
| 187 ASSERT(charactersToAppend); | |
| 188 | |
| 189 unsigned strLength = m_impl->length(); | |
| 190 | |
| 191 if (m_impl->is8Bit()) { | |
| 192 RELEASE_ASSERT(lengthToAppend <= std::numeric_limits<unsigned>::max() -
strLength); | |
| 193 LChar* data; | |
| 194 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(strLength +
lengthToAppend, data); | |
| 195 StringImpl::copyChars(data, m_impl->characters8(), strLength); | |
| 196 StringImpl::copyChars(data + strLength, charactersToAppend, lengthToAppe
nd); | |
| 197 m_impl = newImpl.release(); | |
| 198 return; | |
| 199 } | |
| 200 | |
| 201 RELEASE_ASSERT(lengthToAppend <= std::numeric_limits<unsigned>::max() - strL
ength); | 179 RELEASE_ASSERT(lengthToAppend <= std::numeric_limits<unsigned>::max() - strL
ength); |
| 202 UChar* data; | 180 LChar* data; |
| 203 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(length() + leng
thToAppend, data); | 181 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(strLength + len
gthToAppend, data); |
| 204 StringImpl::copyChars(data, m_impl->characters16(), strLength); | 182 StringImpl::copyChars(data, m_impl->characters8(), strLength); |
| 205 StringImpl::copyChars(data + strLength, charactersToAppend, lengthToAppend); | 183 StringImpl::copyChars(data + strLength, charactersToAppend, lengthToAppend); |
| 206 m_impl = newImpl.release(); | 184 m_impl = newImpl.release(); |
| 207 } | 185 return; |
| 208 | 186 } |
| 209 void String::append(const UChar* charactersToAppend, unsigned lengthToAppend) | 187 |
| 210 { | 188 RELEASE_ASSERT(lengthToAppend <= std::numeric_limits<unsigned>::max() - strLen
gth); |
| 211 if (!m_impl) { | 189 UChar* data; |
| 212 if (!charactersToAppend) | 190 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(length() + length
ToAppend, data); |
| 213 return; | 191 StringImpl::copyChars(data, m_impl->characters16(), strLength); |
| 214 m_impl = StringImpl::create(charactersToAppend, lengthToAppend); | 192 StringImpl::copyChars(data + strLength, charactersToAppend, lengthToAppend); |
| 215 return; | 193 m_impl = newImpl.release(); |
| 216 } | 194 } |
| 217 | 195 |
| 218 if (!lengthToAppend) | 196 void String::append(const UChar* charactersToAppend, unsigned lengthToAppend) { |
| 219 return; | 197 if (!m_impl) { |
| 220 | 198 if (!charactersToAppend) |
| 221 unsigned strLength = m_impl->length(); | 199 return; |
| 222 | 200 m_impl = StringImpl::create(charactersToAppend, lengthToAppend); |
| 223 ASSERT(charactersToAppend); | 201 return; |
| 224 RELEASE_ASSERT(lengthToAppend <= std::numeric_limits<unsigned>::max() - strL
ength); | 202 } |
| 203 |
| 204 if (!lengthToAppend) |
| 205 return; |
| 206 |
| 207 unsigned strLength = m_impl->length(); |
| 208 |
| 209 ASSERT(charactersToAppend); |
| 210 RELEASE_ASSERT(lengthToAppend <= std::numeric_limits<unsigned>::max() - strLen
gth); |
| 211 UChar* data; |
| 212 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(strLength + lengt
hToAppend, data); |
| 213 if (m_impl->is8Bit()) |
| 214 StringImpl::copyChars(data, characters8(), strLength); |
| 215 else |
| 216 StringImpl::copyChars(data, characters16(), strLength); |
| 217 StringImpl::copyChars(data + strLength, charactersToAppend, lengthToAppend); |
| 218 m_impl = newImpl.release(); |
| 219 } |
| 220 |
| 221 template <typename CharType> |
| 222 PassRefPtr<StringImpl> insertInternal(PassRefPtr<StringImpl> impl, const CharTyp
e* charactersToInsert, unsigned lengthToInsert, unsigned position) { |
| 223 if (!lengthToInsert) |
| 224 return impl; |
| 225 |
| 226 ASSERT(charactersToInsert); |
| 227 UChar* data; // FIXME: We should be able to create an 8 bit string here. |
| 228 RELEASE_ASSERT(lengthToInsert <= std::numeric_limits<unsigned>::max() - impl->
length()); |
| 229 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(impl->length() +
lengthToInsert, data); |
| 230 |
| 231 if (impl->is8Bit()) |
| 232 StringImpl::copyChars(data, impl->characters8(), position); |
| 233 else |
| 234 StringImpl::copyChars(data, impl->characters16(), position); |
| 235 |
| 236 StringImpl::copyChars(data + position, charactersToInsert, lengthToInsert); |
| 237 |
| 238 if (impl->is8Bit()) |
| 239 StringImpl::copyChars(data + position + lengthToInsert, impl->characters8()
+ position, impl->length() - position); |
| 240 else |
| 241 StringImpl::copyChars(data + position + lengthToInsert, impl->characters16()
+ position, impl->length() - position); |
| 242 |
| 243 return newImpl.release(); |
| 244 } |
| 245 |
| 246 void String::insert(const UChar* charactersToInsert, unsigned lengthToInsert, un
signed position) { |
| 247 if (position >= length()) { |
| 248 append(charactersToInsert, lengthToInsert); |
| 249 return; |
| 250 } |
| 251 ASSERT(m_impl); |
| 252 m_impl = insertInternal(m_impl.release(), charactersToInsert, lengthToInsert,
position); |
| 253 } |
| 254 |
| 255 void String::insert(const LChar* charactersToInsert, unsigned lengthToInsert, un
signed position) { |
| 256 if (position >= length()) { |
| 257 append(charactersToInsert, lengthToInsert); |
| 258 return; |
| 259 } |
| 260 ASSERT(m_impl); |
| 261 m_impl = insertInternal(m_impl.release(), charactersToInsert, lengthToInsert,
position); |
| 262 } |
| 263 |
| 264 UChar32 String::characterStartingAt(unsigned i) const { |
| 265 if (!m_impl || i >= m_impl->length()) |
| 266 return 0; |
| 267 return m_impl->characterStartingAt(i); |
| 268 } |
| 269 |
| 270 void String::ensure16Bit() { |
| 271 unsigned length = this->length(); |
| 272 if (!length || !is8Bit()) |
| 273 return; |
| 274 m_impl = make16BitFrom8BitSource(m_impl->characters8(), length).impl(); |
| 275 } |
| 276 |
| 277 void String::truncate(unsigned position) { |
| 278 if (position >= length()) |
| 279 return; |
| 280 if (m_impl->is8Bit()) { |
| 281 LChar* data; |
| 282 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(position, data)
; |
| 283 memcpy(data, m_impl->characters8(), position * sizeof(LChar)); |
| 284 m_impl = newImpl.release(); |
| 285 } else { |
| 225 UChar* data; | 286 UChar* data; |
| 226 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(strLength + len
gthToAppend, data); | 287 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(position, data)
; |
| 227 if (m_impl->is8Bit()) | 288 memcpy(data, m_impl->characters16(), position * sizeof(UChar)); |
| 228 StringImpl::copyChars(data, characters8(), strLength); | |
| 229 else | |
| 230 StringImpl::copyChars(data, characters16(), strLength); | |
| 231 StringImpl::copyChars(data + strLength, charactersToAppend, lengthToAppend); | |
| 232 m_impl = newImpl.release(); | 289 m_impl = newImpl.release(); |
| 233 } | 290 } |
| 234 | |
| 235 template<typename CharType> | |
| 236 PassRefPtr<StringImpl> insertInternal(PassRefPtr<StringImpl> impl, const CharTyp
e* charactersToInsert, unsigned lengthToInsert, unsigned position) | |
| 237 { | |
| 238 if (!lengthToInsert) | |
| 239 return impl; | |
| 240 | |
| 241 ASSERT(charactersToInsert); | |
| 242 UChar* data; // FIXME: We should be able to create an 8 bit string here. | |
| 243 RELEASE_ASSERT(lengthToInsert <= std::numeric_limits<unsigned>::max() - impl
->length()); | |
| 244 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(impl->length()
+ lengthToInsert, data); | |
| 245 | |
| 246 if (impl->is8Bit()) | |
| 247 StringImpl::copyChars(data, impl->characters8(), position); | |
| 248 else | |
| 249 StringImpl::copyChars(data, impl->characters16(), position); | |
| 250 | |
| 251 StringImpl::copyChars(data + position, charactersToInsert, lengthToInsert); | |
| 252 | |
| 253 if (impl->is8Bit()) | |
| 254 StringImpl::copyChars(data + position + lengthToInsert, impl->characters
8() + position, impl->length() - position); | |
| 255 else | |
| 256 StringImpl::copyChars(data + position + lengthToInsert, impl->characters
16() + position, impl->length() - position); | |
| 257 | |
| 258 return newImpl.release(); | |
| 259 } | |
| 260 | |
| 261 void String::insert(const UChar* charactersToInsert, unsigned lengthToInsert, un
signed position) | |
| 262 { | |
| 263 if (position >= length()) { | |
| 264 append(charactersToInsert, lengthToInsert); | |
| 265 return; | |
| 266 } | |
| 267 ASSERT(m_impl); | |
| 268 m_impl = insertInternal(m_impl.release(), charactersToInsert, lengthToInsert
, position); | |
| 269 } | |
| 270 | |
| 271 void String::insert(const LChar* charactersToInsert, unsigned lengthToInsert, un
signed position) | |
| 272 { | |
| 273 if (position >= length()) { | |
| 274 append(charactersToInsert, lengthToInsert); | |
| 275 return; | |
| 276 } | |
| 277 ASSERT(m_impl); | |
| 278 m_impl = insertInternal(m_impl.release(), charactersToInsert, lengthToInsert
, position); | |
| 279 } | |
| 280 | |
| 281 UChar32 String::characterStartingAt(unsigned i) const | |
| 282 { | |
| 283 if (!m_impl || i >= m_impl->length()) | |
| 284 return 0; | |
| 285 return m_impl->characterStartingAt(i); | |
| 286 } | |
| 287 | |
| 288 void String::ensure16Bit() | |
| 289 { | |
| 290 unsigned length = this->length(); | |
| 291 if (!length || !is8Bit()) | |
| 292 return; | |
| 293 m_impl = make16BitFrom8BitSource(m_impl->characters8(), length).impl(); | |
| 294 } | |
| 295 | |
| 296 void String::truncate(unsigned position) | |
| 297 { | |
| 298 if (position >= length()) | |
| 299 return; | |
| 300 if (m_impl->is8Bit()) { | |
| 301 LChar* data; | |
| 302 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(position, d
ata); | |
| 303 memcpy(data, m_impl->characters8(), position * sizeof(LChar)); | |
| 304 m_impl = newImpl.release(); | |
| 305 } else { | |
| 306 UChar* data; | |
| 307 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(position, d
ata); | |
| 308 memcpy(data, m_impl->characters16(), position * sizeof(UChar)); | |
| 309 m_impl = newImpl.release(); | |
| 310 } | |
| 311 } | 291 } |
| 312 | 292 |
| 313 template <typename CharacterType> | 293 template <typename CharacterType> |
| 314 inline void String::removeInternal(const CharacterType* characters, unsigned pos
ition, int lengthToRemove) | 294 inline void String::removeInternal(const CharacterType* characters, unsigned pos
ition, int lengthToRemove) { |
| 315 { | 295 CharacterType* data; |
| 316 CharacterType* data; | 296 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(length() - length
ToRemove, data); |
| 317 RefPtr<StringImpl> newImpl = StringImpl::createUninitialized(length() - leng
thToRemove, data); | 297 memcpy(data, characters, position * sizeof(CharacterType)); |
| 318 memcpy(data, characters, position * sizeof(CharacterType)); | 298 memcpy(data + position, characters + position + lengthToRemove, |
| 319 memcpy(data + position, characters + position + lengthToRemove, | 299 (length() - lengthToRemove - position) * sizeof(CharacterType)); |
| 320 (length() - lengthToRemove - position) * sizeof(CharacterType)); | 300 |
| 321 | 301 m_impl = newImpl.release(); |
| 322 m_impl = newImpl.release(); | 302 } |
| 323 } | 303 |
| 324 | 304 void String::remove(unsigned position, int lengthToRemove) { |
| 325 void String::remove(unsigned position, int lengthToRemove) | 305 if (lengthToRemove <= 0) |
| 326 { | 306 return; |
| 327 if (lengthToRemove <= 0) | 307 if (position >= length()) |
| 328 return; | 308 return; |
| 329 if (position >= length()) | 309 if (static_cast<unsigned>(lengthToRemove) > length() - position) |
| 330 return; | 310 lengthToRemove = length() - position; |
| 331 if (static_cast<unsigned>(lengthToRemove) > length() - position) | 311 |
| 332 lengthToRemove = length() - position; | 312 if (is8Bit()) { |
| 333 | 313 removeInternal(characters8(), position, lengthToRemove); |
| 334 if (is8Bit()) { | 314 |
| 335 removeInternal(characters8(), position, lengthToRemove); | 315 return; |
| 336 | 316 } |
| 337 return; | 317 |
| 338 } | 318 removeInternal(characters16(), position, lengthToRemove); |
| 339 | 319 } |
| 340 removeInternal(characters16(), position, lengthToRemove); | 320 |
| 341 } | 321 String String::substring(unsigned pos, unsigned len) const { |
| 342 | 322 if (!m_impl) |
| 343 String String::substring(unsigned pos, unsigned len) const | 323 return String(); |
| 344 { | 324 return m_impl->substring(pos, len); |
| 345 if (!m_impl) | 325 } |
| 346 return String(); | 326 |
| 347 return m_impl->substring(pos, len); | 327 String String::lower() const { |
| 348 } | 328 if (!m_impl) |
| 349 | 329 return String(); |
| 350 String String::lower() const | 330 return m_impl->lower(); |
| 351 { | 331 } |
| 352 if (!m_impl) | 332 |
| 353 return String(); | 333 String String::upper() const { |
| 354 return m_impl->lower(); | 334 if (!m_impl) |
| 355 } | 335 return String(); |
| 356 | 336 return m_impl->upper(); |
| 357 String String::upper() const | 337 } |
| 358 { | 338 |
| 359 if (!m_impl) | 339 String String::lower(const AtomicString& localeIdentifier) const { |
| 360 return String(); | 340 if (!m_impl) |
| 361 return m_impl->upper(); | 341 return String(); |
| 362 } | 342 return m_impl->lower(localeIdentifier); |
| 363 | 343 } |
| 364 String String::lower(const AtomicString& localeIdentifier) const | 344 |
| 365 { | 345 String String::upper(const AtomicString& localeIdentifier) const { |
| 366 if (!m_impl) | 346 if (!m_impl) |
| 367 return String(); | 347 return String(); |
| 368 return m_impl->lower(localeIdentifier); | 348 return m_impl->upper(localeIdentifier); |
| 369 } | 349 } |
| 370 | 350 |
| 371 String String::upper(const AtomicString& localeIdentifier) const | 351 String String::stripWhiteSpace() const { |
| 372 { | 352 if (!m_impl) |
| 373 if (!m_impl) | 353 return String(); |
| 374 return String(); | 354 return m_impl->stripWhiteSpace(); |
| 375 return m_impl->upper(localeIdentifier); | 355 } |
| 376 } | 356 |
| 377 | 357 String String::stripWhiteSpace(IsWhiteSpaceFunctionPtr isWhiteSpace) const { |
| 378 String String::stripWhiteSpace() const | 358 if (!m_impl) |
| 379 { | 359 return String(); |
| 380 if (!m_impl) | 360 return m_impl->stripWhiteSpace(isWhiteSpace); |
| 381 return String(); | 361 } |
| 382 return m_impl->stripWhiteSpace(); | 362 |
| 383 } | 363 String String::simplifyWhiteSpace(StripBehavior stripBehavior) const { |
| 384 | 364 if (!m_impl) |
| 385 String String::stripWhiteSpace(IsWhiteSpaceFunctionPtr isWhiteSpace) const | 365 return String(); |
| 386 { | 366 return m_impl->simplifyWhiteSpace(stripBehavior); |
| 387 if (!m_impl) | 367 } |
| 388 return String(); | 368 |
| 389 return m_impl->stripWhiteSpace(isWhiteSpace); | 369 String String::simplifyWhiteSpace(IsWhiteSpaceFunctionPtr isWhiteSpace, StripBeh
avior stripBehavior) const { |
| 390 } | 370 if (!m_impl) |
| 391 | 371 return String(); |
| 392 String String::simplifyWhiteSpace(StripBehavior stripBehavior) const | 372 return m_impl->simplifyWhiteSpace(isWhiteSpace, stripBehavior); |
| 393 { | 373 } |
| 394 if (!m_impl) | 374 |
| 395 return String(); | 375 String String::removeCharacters(CharacterMatchFunctionPtr findMatch) const { |
| 396 return m_impl->simplifyWhiteSpace(stripBehavior); | 376 if (!m_impl) |
| 397 } | 377 return String(); |
| 398 | 378 return m_impl->removeCharacters(findMatch); |
| 399 String String::simplifyWhiteSpace(IsWhiteSpaceFunctionPtr isWhiteSpace, StripBeh
avior stripBehavior) const | 379 } |
| 400 { | 380 |
| 401 if (!m_impl) | 381 String String::foldCase() const { |
| 402 return String(); | 382 if (!m_impl) |
| 403 return m_impl->simplifyWhiteSpace(isWhiteSpace, stripBehavior); | 383 return String(); |
| 404 } | 384 return m_impl->foldCase(); |
| 405 | 385 } |
| 406 String String::removeCharacters(CharacterMatchFunctionPtr findMatch) const | 386 |
| 407 { | 387 Vector<UChar> String::charactersWithNullTermination() const { |
| 408 if (!m_impl) | 388 if (!m_impl) |
| 409 return String(); | 389 return Vector<UChar>(); |
| 410 return m_impl->removeCharacters(findMatch); | 390 |
| 411 } | 391 Vector<UChar> result; |
| 412 | 392 result.reserveInitialCapacity(length() + 1); |
| 413 String String::foldCase() const | 393 appendTo(result); |
| 414 { | 394 result.append('\0'); |
| 415 if (!m_impl) | 395 return result; |
| 416 return String(); | 396 } |
| 417 return m_impl->foldCase(); | 397 |
| 418 } | 398 unsigned String::copyTo(UChar* buffer, unsigned pos, unsigned maxLength) const { |
| 419 | 399 unsigned length = this->length(); |
| 420 Vector<UChar> String::charactersWithNullTermination() const | 400 RELEASE_ASSERT(pos <= length); |
| 421 { | 401 unsigned numCharacters = std::min(length - pos, maxLength); |
| 422 if (!m_impl) | 402 if (!numCharacters) |
| 423 return Vector<UChar>(); | 403 return 0; |
| 424 | 404 if (is8Bit()) |
| 425 Vector<UChar> result; | 405 StringImpl::copyChars(buffer, characters8() + pos, numCharacters); |
| 426 result.reserveInitialCapacity(length() + 1); | 406 else |
| 427 appendTo(result); | 407 StringImpl::copyChars(buffer, characters16() + pos, numCharacters); |
| 428 result.append('\0'); | 408 return numCharacters; |
| 429 return result; | 409 } |
| 430 } | 410 |
| 431 | 411 String String::format(const char* format, ...) { |
| 432 unsigned String::copyTo(UChar* buffer, unsigned pos, unsigned maxLength) const | 412 va_list args; |
| 433 { | 413 va_start(args, format); |
| 434 unsigned length = this->length(); | 414 |
| 435 RELEASE_ASSERT(pos <= length); | 415 // Do the format once to get the length. |
| 436 unsigned numCharacters = std::min(length - pos, maxLength); | |
| 437 if (!numCharacters) | |
| 438 return 0; | |
| 439 if (is8Bit()) | |
| 440 StringImpl::copyChars(buffer, characters8() + pos, numCharacters); | |
| 441 else | |
| 442 StringImpl::copyChars(buffer, characters16() + pos, numCharacters); | |
| 443 return numCharacters; | |
| 444 } | |
| 445 | |
| 446 String String::format(const char *format, ...) | |
| 447 { | |
| 448 va_list args; | |
| 449 va_start(args, format); | |
| 450 | |
| 451 // Do the format once to get the length. | |
| 452 #if COMPILER(MSVC) | 416 #if COMPILER(MSVC) |
| 453 int result = _vscprintf(format, args); | 417 int result = _vscprintf(format, args); |
| 454 #else | 418 #else |
| 455 char ch; | 419 char ch; |
| 456 int result = vsnprintf(&ch, 1, format, args); | 420 int result = vsnprintf(&ch, 1, format, args); |
| 457 // We need to call va_end() and then va_start() again here, as the | 421 // We need to call va_end() and then va_start() again here, as the |
| 458 // contents of args is undefined after the call to vsnprintf | 422 // contents of args is undefined after the call to vsnprintf |
| 459 // according to http://man.cx/snprintf(3) | 423 // according to http://man.cx/snprintf(3) |
| 460 // | 424 // |
| 461 // Not calling va_end/va_start here happens to work on lots of | 425 // Not calling va_end/va_start here happens to work on lots of |
| 462 // systems, but fails e.g. on 64bit Linux. | 426 // systems, but fails e.g. on 64bit Linux. |
| 463 #endif | 427 #endif |
| 464 va_end(args); | 428 va_end(args); |
| 465 | 429 |
| 466 if (result == 0) | 430 if (result == 0) |
| 467 return String(""); | 431 return String(""); |
| 468 if (result < 0) | 432 if (result < 0) |
| 469 return String(); | 433 return String(); |
| 470 | 434 |
| 471 Vector<char, 256> buffer; | 435 Vector<char, 256> buffer; |
| 472 unsigned len = result; | 436 unsigned len = result; |
| 473 buffer.grow(len + 1); | 437 buffer.grow(len + 1); |
| 474 | 438 |
| 475 va_start(args, format); | 439 va_start(args, format); |
| 476 // Now do the formatting again, guaranteed to fit. | 440 // Now do the formatting again, guaranteed to fit. |
| 477 vsnprintf(buffer.data(), buffer.size(), format, args); | 441 vsnprintf(buffer.data(), buffer.size(), format, args); |
| 478 | 442 |
| 479 va_end(args); | 443 va_end(args); |
| 480 | 444 |
| 481 return StringImpl::create(reinterpret_cast<const LChar*>(buffer.data()), len
); | 445 return StringImpl::create(reinterpret_cast<const LChar*>(buffer.data()), len); |
| 482 } | 446 } |
| 483 | 447 |
| 484 String String::number(int number) | 448 String String::number(int number) { |
| 485 { | 449 return numberToStringSigned<String>(number); |
| 486 return numberToStringSigned<String>(number); | 450 } |
| 487 } | 451 |
| 488 | 452 String String::number(unsigned number) { |
| 489 String String::number(unsigned number) | 453 return numberToStringUnsigned<String>(number); |
| 490 { | 454 } |
| 491 return numberToStringUnsigned<String>(number); | 455 |
| 492 } | 456 String String::number(long number) { |
| 493 | 457 return numberToStringSigned<String>(number); |
| 494 String String::number(long number) | 458 } |
| 495 { | 459 |
| 496 return numberToStringSigned<String>(number); | 460 String String::number(unsigned long number) { |
| 497 } | 461 return numberToStringUnsigned<String>(number); |
| 498 | 462 } |
| 499 String String::number(unsigned long number) | 463 |
| 500 { | 464 String String::number(long long number) { |
| 501 return numberToStringUnsigned<String>(number); | 465 return numberToStringSigned<String>(number); |
| 502 } | 466 } |
| 503 | 467 |
| 504 String String::number(long long number) | 468 String String::number(unsigned long long number) { |
| 505 { | 469 return numberToStringUnsigned<String>(number); |
| 506 return numberToStringSigned<String>(number); | 470 } |
| 507 } | 471 |
| 508 | 472 String String::number(double number, unsigned precision, TrailingZerosTruncating
Policy trailingZerosTruncatingPolicy) { |
| 509 String String::number(unsigned long long number) | 473 NumberToStringBuffer buffer; |
| 510 { | 474 return String(numberToFixedPrecisionString(number, precision, buffer, trailing
ZerosTruncatingPolicy == TruncateTrailingZeros)); |
| 511 return numberToStringUnsigned<String>(number); | 475 } |
| 512 } | 476 |
| 513 | 477 String String::numberToStringECMAScript(double number) { |
| 514 String String::number(double number, unsigned precision, TrailingZerosTruncating
Policy trailingZerosTruncatingPolicy) | 478 NumberToStringBuffer buffer; |
| 515 { | 479 return String(numberToString(number, buffer)); |
| 516 NumberToStringBuffer buffer; | 480 } |
| 517 return String(numberToFixedPrecisionString(number, precision, buffer, traili
ngZerosTruncatingPolicy == TruncateTrailingZeros)); | 481 |
| 518 } | 482 String String::numberToStringFixedWidth(double number, unsigned decimalPlaces) { |
| 519 | 483 NumberToStringBuffer buffer; |
| 520 String String::numberToStringECMAScript(double number) | 484 return String(numberToFixedWidthString(number, decimalPlaces, buffer)); |
| 521 { | 485 } |
| 522 NumberToStringBuffer buffer; | 486 |
| 523 return String(numberToString(number, buffer)); | 487 int String::toIntStrict(bool* ok, int base) const { |
| 524 } | 488 if (!m_impl) { |
| 525 | 489 if (ok) |
| 526 String String::numberToStringFixedWidth(double number, unsigned decimalPlaces) | 490 *ok = false; |
| 527 { | 491 return 0; |
| 528 NumberToStringBuffer buffer; | 492 } |
| 529 return String(numberToFixedWidthString(number, decimalPlaces, buffer)); | 493 return m_impl->toIntStrict(ok, base); |
| 530 } | 494 } |
| 531 | 495 |
| 532 int String::toIntStrict(bool* ok, int base) const | 496 unsigned String::toUIntStrict(bool* ok, int base) const { |
| 533 { | 497 if (!m_impl) { |
| 534 if (!m_impl) { | 498 if (ok) |
| 535 if (ok) | 499 *ok = false; |
| 536 *ok = false; | 500 return 0; |
| 537 return 0; | 501 } |
| 538 } | 502 return m_impl->toUIntStrict(ok, base); |
| 539 return m_impl->toIntStrict(ok, base); | 503 } |
| 540 } | 504 |
| 541 | 505 int64_t String::toInt64Strict(bool* ok, int base) const { |
| 542 unsigned String::toUIntStrict(bool* ok, int base) const | 506 if (!m_impl) { |
| 543 { | 507 if (ok) |
| 544 if (!m_impl) { | 508 *ok = false; |
| 545 if (ok) | 509 return 0; |
| 546 *ok = false; | 510 } |
| 547 return 0; | 511 return m_impl->toInt64Strict(ok, base); |
| 548 } | 512 } |
| 549 return m_impl->toUIntStrict(ok, base); | 513 |
| 550 } | 514 uint64_t String::toUInt64Strict(bool* ok, int base) const { |
| 551 | 515 if (!m_impl) { |
| 552 int64_t String::toInt64Strict(bool* ok, int base) const | 516 if (ok) |
| 553 { | 517 *ok = false; |
| 554 if (!m_impl) { | 518 return 0; |
| 555 if (ok) | 519 } |
| 556 *ok = false; | 520 return m_impl->toUInt64Strict(ok, base); |
| 557 return 0; | 521 } |
| 558 } | 522 |
| 559 return m_impl->toInt64Strict(ok, base); | 523 int String::toInt(bool* ok) const { |
| 560 } | 524 if (!m_impl) { |
| 561 | 525 if (ok) |
| 562 uint64_t String::toUInt64Strict(bool* ok, int base) const | 526 *ok = false; |
| 563 { | 527 return 0; |
| 564 if (!m_impl) { | 528 } |
| 565 if (ok) | 529 return m_impl->toInt(ok); |
| 566 *ok = false; | 530 } |
| 567 return 0; | 531 |
| 568 } | 532 unsigned String::toUInt(bool* ok) const { |
| 569 return m_impl->toUInt64Strict(ok, base); | 533 if (!m_impl) { |
| 570 } | 534 if (ok) |
| 571 | 535 *ok = false; |
| 572 int String::toInt(bool* ok) const | 536 return 0; |
| 573 { | 537 } |
| 574 if (!m_impl) { | 538 return m_impl->toUInt(ok); |
| 575 if (ok) | 539 } |
| 576 *ok = false; | 540 |
| 577 return 0; | 541 int64_t String::toInt64(bool* ok) const { |
| 578 } | 542 if (!m_impl) { |
| 579 return m_impl->toInt(ok); | 543 if (ok) |
| 580 } | 544 *ok = false; |
| 581 | 545 return 0; |
| 582 unsigned String::toUInt(bool* ok) const | 546 } |
| 583 { | 547 return m_impl->toInt64(ok); |
| 584 if (!m_impl) { | 548 } |
| 585 if (ok) | 549 |
| 586 *ok = false; | 550 uint64_t String::toUInt64(bool* ok) const { |
| 587 return 0; | 551 if (!m_impl) { |
| 588 } | 552 if (ok) |
| 589 return m_impl->toUInt(ok); | 553 *ok = false; |
| 590 } | 554 return 0; |
| 591 | 555 } |
| 592 int64_t String::toInt64(bool* ok) const | 556 return m_impl->toUInt64(ok); |
| 593 { | 557 } |
| 594 if (!m_impl) { | 558 |
| 595 if (ok) | 559 double String::toDouble(bool* ok) const { |
| 596 *ok = false; | 560 if (!m_impl) { |
| 597 return 0; | 561 if (ok) |
| 598 } | 562 *ok = false; |
| 599 return m_impl->toInt64(ok); | 563 return 0.0; |
| 600 } | 564 } |
| 601 | 565 return m_impl->toDouble(ok); |
| 602 uint64_t String::toUInt64(bool* ok) const | 566 } |
| 603 { | 567 |
| 604 if (!m_impl) { | 568 float String::toFloat(bool* ok) const { |
| 605 if (ok) | 569 if (!m_impl) { |
| 606 *ok = false; | 570 if (ok) |
| 607 return 0; | 571 *ok = false; |
| 608 } | 572 return 0.0f; |
| 609 return m_impl->toUInt64(ok); | 573 } |
| 610 } | 574 return m_impl->toFloat(ok); |
| 611 | 575 } |
| 612 double String::toDouble(bool* ok) const | 576 |
| 613 { | 577 String String::isolatedCopy() const { |
| 614 if (!m_impl) { | 578 if (!m_impl) |
| 615 if (ok) | 579 return String(); |
| 616 *ok = false; | 580 return m_impl->isolatedCopy(); |
| 617 return 0.0; | 581 } |
| 618 } | 582 |
| 619 return m_impl->toDouble(ok); | 583 bool String::isSafeToSendToAnotherThread() const { |
| 620 } | 584 if (!impl()) |
| 621 | 585 return true; |
| 622 float String::toFloat(bool* ok) const | 586 if (impl()->isStatic()) |
| 623 { | 587 return true; |
| 624 if (!m_impl) { | 588 // AtomicStrings are not safe to send between threads as ~StringImpl() |
| 625 if (ok) | 589 // will try to remove them from the wrong AtomicStringTable. |
| 626 *ok = false; | 590 if (impl()->isAtomic()) |
| 627 return 0.0f; | |
| 628 } | |
| 629 return m_impl->toFloat(ok); | |
| 630 } | |
| 631 | |
| 632 String String::isolatedCopy() const | |
| 633 { | |
| 634 if (!m_impl) | |
| 635 return String(); | |
| 636 return m_impl->isolatedCopy(); | |
| 637 } | |
| 638 | |
| 639 bool String::isSafeToSendToAnotherThread() const | |
| 640 { | |
| 641 if (!impl()) | |
| 642 return true; | |
| 643 if (impl()->isStatic()) | |
| 644 return true; | |
| 645 // AtomicStrings are not safe to send between threads as ~StringImpl() | |
| 646 // will try to remove them from the wrong AtomicStringTable. | |
| 647 if (impl()->isAtomic()) | |
| 648 return false; | |
| 649 if (impl()->hasOneRef()) | |
| 650 return true; | |
| 651 return false; | 591 return false; |
| 652 } | 592 if (impl()->hasOneRef()) |
| 653 | 593 return true; |
| 654 void String::split(const String& separator, bool allowEmptyEntries, Vector<Strin
g>& result) const | 594 return false; |
| 655 { | 595 } |
| 656 result.clear(); | 596 |
| 657 | 597 void String::split(const String& separator, bool allowEmptyEntries, Vector<Strin
g>& result) const { |
| 658 unsigned startPos = 0; | 598 result.clear(); |
| 659 size_t endPos; | 599 |
| 660 while ((endPos = find(separator, startPos)) != kNotFound) { | 600 unsigned startPos = 0; |
| 661 if (allowEmptyEntries || startPos != endPos) | 601 size_t endPos; |
| 662 result.append(substring(startPos, endPos - startPos)); | 602 while ((endPos = find(separator, startPos)) != kNotFound) { |
| 663 startPos = endPos + separator.length(); | 603 if (allowEmptyEntries || startPos != endPos) |
| 664 } | 604 result.append(substring(startPos, endPos - startPos)); |
| 665 if (allowEmptyEntries || startPos != length()) | 605 startPos = endPos + separator.length(); |
| 666 result.append(substring(startPos)); | 606 } |
| 667 } | 607 if (allowEmptyEntries || startPos != length()) |
| 668 | 608 result.append(substring(startPos)); |
| 669 void String::split(UChar separator, bool allowEmptyEntries, Vector<String>& resu
lt) const | 609 } |
| 670 { | 610 |
| 671 result.clear(); | 611 void String::split(UChar separator, bool allowEmptyEntries, Vector<String>& resu
lt) const { |
| 672 | 612 result.clear(); |
| 673 unsigned startPos = 0; | 613 |
| 674 size_t endPos; | 614 unsigned startPos = 0; |
| 675 while ((endPos = find(separator, startPos)) != kNotFound) { | 615 size_t endPos; |
| 676 if (allowEmptyEntries || startPos != endPos) | 616 while ((endPos = find(separator, startPos)) != kNotFound) { |
| 677 result.append(substring(startPos, endPos - startPos)); | 617 if (allowEmptyEntries || startPos != endPos) |
| 678 startPos = endPos + 1; | 618 result.append(substring(startPos, endPos - startPos)); |
| 679 } | 619 startPos = endPos + 1; |
| 680 if (allowEmptyEntries || startPos != length()) | 620 } |
| 681 result.append(substring(startPos)); | 621 if (allowEmptyEntries || startPos != length()) |
| 682 } | 622 result.append(substring(startPos)); |
| 683 | 623 } |
| 684 CString String::ascii() const | 624 |
| 685 { | 625 CString String::ascii() const { |
| 686 // Printable ASCII characters 32..127 and the null character are | 626 // Printable ASCII characters 32..127 and the null character are |
| 687 // preserved, characters outside of this range are converted to '?'. | 627 // preserved, characters outside of this range are converted to '?'. |
| 688 | 628 |
| 689 unsigned length = this->length(); | 629 unsigned length = this->length(); |
| 690 if (!length) { | 630 if (!length) { |
| 691 char* characterBuffer; | 631 char* characterBuffer; |
| 692 return CString::newUninitialized(length, characterBuffer); | 632 return CString::newUninitialized(length, characterBuffer); |
| 693 } | 633 } |
| 694 | 634 |
| 695 if (this->is8Bit()) { | 635 if (this->is8Bit()) { |
| 696 const LChar* characters = this->characters8(); | 636 const LChar* characters = this->characters8(); |
| 697 | |
| 698 char* characterBuffer; | |
| 699 CString result = CString::newUninitialized(length, characterBuffer); | |
| 700 | |
| 701 for (unsigned i = 0; i < length; ++i) { | |
| 702 LChar ch = characters[i]; | |
| 703 characterBuffer[i] = ch && (ch < 0x20 || ch > 0x7f) ? '?' : ch; | |
| 704 } | |
| 705 | |
| 706 return result; | |
| 707 } | |
| 708 | |
| 709 const UChar* characters = this->characters16(); | |
| 710 | 637 |
| 711 char* characterBuffer; | 638 char* characterBuffer; |
| 712 CString result = CString::newUninitialized(length, characterBuffer); | 639 CString result = CString::newUninitialized(length, characterBuffer); |
| 713 | 640 |
| 714 for (unsigned i = 0; i < length; ++i) { | 641 for (unsigned i = 0; i < length; ++i) { |
| 715 UChar ch = characters[i]; | 642 LChar ch = characters[i]; |
| 716 characterBuffer[i] = ch && (ch < 0x20 || ch > 0x7f) ? '?' : static_cast<
char>(ch); | 643 characterBuffer[i] = ch && (ch < 0x20 || ch > 0x7f) ? '?' : ch; |
| 717 } | 644 } |
| 718 | 645 |
| 719 return result; | 646 return result; |
| 720 } | 647 } |
| 721 | 648 |
| 722 CString String::latin1() const | 649 const UChar* characters = this->characters16(); |
| 723 { | 650 |
| 724 // Basic Latin1 (ISO) encoding - Unicode characters 0..255 are | 651 char* characterBuffer; |
| 725 // preserved, characters outside of this range are converted to '?'. | 652 CString result = CString::newUninitialized(length, characterBuffer); |
| 726 | 653 |
| 727 unsigned length = this->length(); | 654 for (unsigned i = 0; i < length; ++i) { |
| 728 | 655 UChar ch = characters[i]; |
| 729 if (!length) | 656 characterBuffer[i] = ch && (ch < 0x20 || ch > 0x7f) ? '?' : static_cast<char
>(ch); |
| 730 return CString("", 0); | 657 } |
| 731 | 658 |
| 732 if (is8Bit()) | 659 return result; |
| 733 return CString(reinterpret_cast<const char*>(this->characters8()), lengt
h); | 660 } |
| 734 | 661 |
| 662 CString String::latin1() const { |
| 663 // Basic Latin1 (ISO) encoding - Unicode characters 0..255 are |
| 664 // preserved, characters outside of this range are converted to '?'. |
| 665 |
| 666 unsigned length = this->length(); |
| 667 |
| 668 if (!length) |
| 669 return CString("", 0); |
| 670 |
| 671 if (is8Bit()) |
| 672 return CString(reinterpret_cast<const char*>(this->characters8()), length); |
| 673 |
| 674 const UChar* characters = this->characters16(); |
| 675 |
| 676 char* characterBuffer; |
| 677 CString result = CString::newUninitialized(length, characterBuffer); |
| 678 |
| 679 for (unsigned i = 0; i < length; ++i) { |
| 680 UChar ch = characters[i]; |
| 681 characterBuffer[i] = ch > 0xff ? '?' : static_cast<char>(ch); |
| 682 } |
| 683 |
| 684 return result; |
| 685 } |
| 686 |
| 687 // Helper to write a three-byte UTF-8 code point to the buffer, caller must chec
k room is available. |
| 688 static inline void putUTF8Triple(char*& buffer, UChar ch) { |
| 689 ASSERT(ch >= 0x0800); |
| 690 *buffer++ = static_cast<char>(((ch >> 12) & 0x0F) | 0xE0); |
| 691 *buffer++ = static_cast<char>(((ch >> 6) & 0x3F) | 0x80); |
| 692 *buffer++ = static_cast<char>((ch & 0x3F) | 0x80); |
| 693 } |
| 694 |
| 695 CString String::utf8(UTF8ConversionMode mode) const { |
| 696 unsigned length = this->length(); |
| 697 |
| 698 if (!length) |
| 699 return CString("", 0); |
| 700 |
| 701 // Allocate a buffer big enough to hold all the characters |
| 702 // (an individual UTF-16 UChar can only expand to 3 UTF-8 bytes). |
| 703 // Optimization ideas, if we find this function is hot: |
| 704 // * We could speculatively create a CStringBuffer to contain 'length' |
| 705 // characters, and resize if necessary (i.e. if the buffer contains |
| 706 // non-ascii characters). (Alternatively, scan the buffer first for |
| 707 // ascii characters, so we know this will be sufficient). |
| 708 // * We could allocate a CStringBuffer with an appropriate size to |
| 709 // have a good chance of being able to write the string into the |
| 710 // buffer without reallocing (say, 1.5 x length). |
| 711 if (length > std::numeric_limits<unsigned>::max() / 3) |
| 712 return CString(); |
| 713 Vector<char, 1024> bufferVector(length * 3); |
| 714 |
| 715 char* buffer = bufferVector.data(); |
| 716 |
| 717 if (is8Bit()) { |
| 718 const LChar* characters = this->characters8(); |
| 719 |
| 720 ConversionResult result = convertLatin1ToUTF8(&characters, characters + leng
th, &buffer, buffer + bufferVector.size()); |
| 721 ASSERT_UNUSED(result, result != targetExhausted); // (length * 3) should be
sufficient for any conversion |
| 722 } else { |
| 735 const UChar* characters = this->characters16(); | 723 const UChar* characters = this->characters16(); |
| 736 | 724 |
| 737 char* characterBuffer; | 725 if (mode == StrictUTF8ConversionReplacingUnpairedSurrogatesWithFFFD) { |
| 738 CString result = CString::newUninitialized(length, characterBuffer); | 726 const UChar* charactersEnd = characters + length; |
| 739 | 727 char* bufferEnd = buffer + bufferVector.size(); |
| 740 for (unsigned i = 0; i < length; ++i) { | 728 while (characters < charactersEnd) { |
| 741 UChar ch = characters[i]; | 729 // Use strict conversion to detect unpaired surrogates. |
| 742 characterBuffer[i] = ch > 0xff ? '?' : static_cast<char>(ch); | 730 ConversionResult result = convertUTF16ToUTF8(&characters, charactersEnd,
&buffer, bufferEnd, true); |
| 731 ASSERT(result != targetExhausted); |
| 732 // Conversion fails when there is an unpaired surrogate. Put |
| 733 // replacement character (U+FFFD) instead of the unpaired |
| 734 // surrogate. |
| 735 if (result != conversionOK) { |
| 736 ASSERT((0xD800 <= *characters && *characters <= 0xDFFF)); |
| 737 // There should be room left, since one UChar hasn't been |
| 738 // converted. |
| 739 ASSERT((buffer + 3) <= bufferEnd); |
| 740 putUTF8Triple(buffer, replacementCharacter); |
| 741 ++characters; |
| 742 } |
| 743 } |
| 744 } else { |
| 745 bool strict = mode == StrictUTF8Conversion; |
| 746 ConversionResult result = convertUTF16ToUTF8(&characters, characters + len
gth, &buffer, buffer + bufferVector.size(), strict); |
| 747 ASSERT(result != targetExhausted); // (length * 3) should be sufficient f
or any conversion |
| 748 |
| 749 // Only produced from strict conversion. |
| 750 if (result == sourceIllegal) { |
| 751 ASSERT(strict); |
| 752 return CString(); |
| 753 } |
| 754 |
| 755 // Check for an unconverted high surrogate. |
| 756 if (result == sourceExhausted) { |
| 757 if (strict) |
| 758 return CString(); |
| 759 // This should be one unpaired high surrogate. Treat it the same |
| 760 // was as an unpaired high surrogate would have been handled in |
| 761 // the middle of a string with non-strict conversion - which is |
| 762 // to say, simply encode it to UTF-8. |
| 763 ASSERT((characters + 1) == (this->characters16() + length)); |
| 764 ASSERT((*characters >= 0xD800) && (*characters <= 0xDBFF)); |
| 765 // There should be room left, since one UChar hasn't been |
| 766 // converted. |
| 767 ASSERT((buffer + 3) <= (buffer + bufferVector.size())); |
| 768 putUTF8Triple(buffer, *characters); |
| 769 } |
| 743 } | 770 } |
| 744 | 771 } |
| 745 return result; | 772 |
| 746 } | 773 return CString(bufferVector.data(), buffer - bufferVector.data()); |
| 747 | 774 } |
| 748 // Helper to write a three-byte UTF-8 code point to the buffer, caller must chec
k room is available. | 775 |
| 749 static inline void putUTF8Triple(char*& buffer, UChar ch) | 776 String String::make8BitFrom16BitSource(const UChar* source, size_t length) { |
| 750 { | 777 if (!length) |
| 751 ASSERT(ch >= 0x0800); | 778 return emptyString(); |
| 752 *buffer++ = static_cast<char>(((ch >> 12) & 0x0F) | 0xE0); | 779 |
| 753 *buffer++ = static_cast<char>(((ch >> 6) & 0x3F) | 0x80); | 780 LChar* destination; |
| 754 *buffer++ = static_cast<char>((ch & 0x3F) | 0x80); | 781 String result = String::createUninitialized(length, destination); |
| 755 } | 782 |
| 756 | 783 copyLCharsFromUCharSource(destination, source, length); |
| 757 CString String::utf8(UTF8ConversionMode mode) const | 784 |
| 758 { | 785 return result; |
| 759 unsigned length = this->length(); | 786 } |
| 760 | 787 |
| 761 if (!length) | 788 String String::make16BitFrom8BitSource(const LChar* source, size_t length) { |
| 762 return CString("", 0); | 789 if (!length) |
| 763 | 790 return emptyString16Bit(); |
| 764 // Allocate a buffer big enough to hold all the characters | 791 |
| 765 // (an individual UTF-16 UChar can only expand to 3 UTF-8 bytes). | 792 UChar* destination; |
| 766 // Optimization ideas, if we find this function is hot: | 793 String result = String::createUninitialized(length, destination); |
| 767 // * We could speculatively create a CStringBuffer to contain 'length' | 794 |
| 768 // characters, and resize if necessary (i.e. if the buffer contains | 795 StringImpl::copyChars(destination, source, length); |
| 769 // non-ascii characters). (Alternatively, scan the buffer first for | 796 |
| 770 // ascii characters, so we know this will be sufficient). | 797 return result; |
| 771 // * We could allocate a CStringBuffer with an appropriate size to | 798 } |
| 772 // have a good chance of being able to write the string into the | 799 |
| 773 // buffer without reallocing (say, 1.5 x length). | 800 String String::fromUTF8(const LChar* stringStart, size_t length) { |
| 774 if (length > std::numeric_limits<unsigned>::max() / 3) | 801 RELEASE_ASSERT(length <= std::numeric_limits<unsigned>::max()); |
| 775 return CString(); | 802 |
| 776 Vector<char, 1024> bufferVector(length * 3); | 803 if (!stringStart) |
| 777 | 804 return String(); |
| 778 char* buffer = bufferVector.data(); | 805 |
| 779 | 806 if (!length) |
| 780 if (is8Bit()) { | 807 return emptyString(); |
| 781 const LChar* characters = this->characters8(); | 808 |
| 782 | 809 if (charactersAreAllASCII(stringStart, length)) |
| 783 ConversionResult result = convertLatin1ToUTF8(&characters, characters +
length, &buffer, buffer + bufferVector.size()); | 810 return StringImpl::create(stringStart, length); |
| 784 ASSERT_UNUSED(result, result != targetExhausted); // (length * 3) should
be sufficient for any conversion | 811 |
| 785 } else { | 812 Vector<UChar, 1024> buffer(length); |
| 786 const UChar* characters = this->characters16(); | 813 UChar* bufferStart = buffer.data(); |
| 787 | 814 |
| 788 if (mode == StrictUTF8ConversionReplacingUnpairedSurrogatesWithFFFD) { | 815 UChar* bufferCurrent = bufferStart; |
| 789 const UChar* charactersEnd = characters + length; | 816 const char* stringCurrent = reinterpret_cast<const char*>(stringStart); |
| 790 char* bufferEnd = buffer + bufferVector.size(); | 817 if (convertUTF8ToUTF16(&stringCurrent, reinterpret_cast<const char*>(stringSta
rt + length), &bufferCurrent, bufferCurrent + buffer.size()) != conversionOK) |
| 791 while (characters < charactersEnd) { | 818 return String(); |
| 792 // Use strict conversion to detect unpaired surrogates. | 819 |
| 793 ConversionResult result = convertUTF16ToUTF8(&characters, charac
tersEnd, &buffer, bufferEnd, true); | 820 unsigned utf16Length = bufferCurrent - bufferStart; |
| 794 ASSERT(result != targetExhausted); | 821 ASSERT(utf16Length < length); |
| 795 // Conversion fails when there is an unpaired surrogate. Put | 822 return StringImpl::create(bufferStart, utf16Length); |
| 796 // replacement character (U+FFFD) instead of the unpaired | 823 } |
| 797 // surrogate. | 824 |
| 798 if (result != conversionOK) { | 825 String String::fromUTF8(const LChar* string) { |
| 799 ASSERT((0xD800 <= *characters && *characters <= 0xDFFF)); | 826 if (!string) |
| 800 // There should be room left, since one UChar hasn't been | 827 return String(); |
| 801 // converted. | 828 return fromUTF8(string, strlen(reinterpret_cast<const char*>(string))); |
| 802 ASSERT((buffer + 3) <= bufferEnd); | 829 } |
| 803 putUTF8Triple(buffer, replacementCharacter); | 830 |
| 804 ++characters; | 831 String String::fromUTF8(const CString& s) { |
| 805 } | 832 return fromUTF8(s.data()); |
| 806 } | 833 } |
| 807 } else { | 834 |
| 808 bool strict = mode == StrictUTF8Conversion; | 835 String String::fromUTF8WithLatin1Fallback(const LChar* string, size_t size) { |
| 809 ConversionResult result = convertUTF16ToUTF8(&characters, characters
+ length, &buffer, buffer + bufferVector.size(), strict); | 836 String utf8 = fromUTF8(string, size); |
| 810 ASSERT(result != targetExhausted); // (length * 3) should be suffici
ent for any conversion | 837 if (!utf8) |
| 811 | 838 return String(string, size); |
| 812 // Only produced from strict conversion. | 839 return utf8; |
| 813 if (result == sourceIllegal) { | |
| 814 ASSERT(strict); | |
| 815 return CString(); | |
| 816 } | |
| 817 | |
| 818 // Check for an unconverted high surrogate. | |
| 819 if (result == sourceExhausted) { | |
| 820 if (strict) | |
| 821 return CString(); | |
| 822 // This should be one unpaired high surrogate. Treat it the same | |
| 823 // was as an unpaired high surrogate would have been handled in | |
| 824 // the middle of a string with non-strict conversion - which is | |
| 825 // to say, simply encode it to UTF-8. | |
| 826 ASSERT((characters + 1) == (this->characters16() + length)); | |
| 827 ASSERT((*characters >= 0xD800) && (*characters <= 0xDBFF)); | |
| 828 // There should be room left, since one UChar hasn't been | |
| 829 // converted. | |
| 830 ASSERT((buffer + 3) <= (buffer + bufferVector.size())); | |
| 831 putUTF8Triple(buffer, *characters); | |
| 832 } | |
| 833 } | |
| 834 } | |
| 835 | |
| 836 return CString(bufferVector.data(), buffer - bufferVector.data()); | |
| 837 } | |
| 838 | |
| 839 String String::make8BitFrom16BitSource(const UChar* source, size_t length) | |
| 840 { | |
| 841 if (!length) | |
| 842 return emptyString(); | |
| 843 | |
| 844 LChar* destination; | |
| 845 String result = String::createUninitialized(length, destination); | |
| 846 | |
| 847 copyLCharsFromUCharSource(destination, source, length); | |
| 848 | |
| 849 return result; | |
| 850 } | |
| 851 | |
| 852 String String::make16BitFrom8BitSource(const LChar* source, size_t length) | |
| 853 { | |
| 854 if (!length) | |
| 855 return emptyString16Bit(); | |
| 856 | |
| 857 UChar* destination; | |
| 858 String result = String::createUninitialized(length, destination); | |
| 859 | |
| 860 StringImpl::copyChars(destination, source, length); | |
| 861 | |
| 862 return result; | |
| 863 } | |
| 864 | |
| 865 String String::fromUTF8(const LChar* stringStart, size_t length) | |
| 866 { | |
| 867 RELEASE_ASSERT(length <= std::numeric_limits<unsigned>::max()); | |
| 868 | |
| 869 if (!stringStart) | |
| 870 return String(); | |
| 871 | |
| 872 if (!length) | |
| 873 return emptyString(); | |
| 874 | |
| 875 if (charactersAreAllASCII(stringStart, length)) | |
| 876 return StringImpl::create(stringStart, length); | |
| 877 | |
| 878 Vector<UChar, 1024> buffer(length); | |
| 879 UChar* bufferStart = buffer.data(); | |
| 880 | |
| 881 UChar* bufferCurrent = bufferStart; | |
| 882 const char* stringCurrent = reinterpret_cast<const char*>(stringStart); | |
| 883 if (convertUTF8ToUTF16(&stringCurrent, reinterpret_cast<const char *>(string
Start + length), &bufferCurrent, bufferCurrent + buffer.size()) != conversionOK) | |
| 884 return String(); | |
| 885 | |
| 886 unsigned utf16Length = bufferCurrent - bufferStart; | |
| 887 ASSERT(utf16Length < length); | |
| 888 return StringImpl::create(bufferStart, utf16Length); | |
| 889 } | |
| 890 | |
| 891 String String::fromUTF8(const LChar* string) | |
| 892 { | |
| 893 if (!string) | |
| 894 return String(); | |
| 895 return fromUTF8(string, strlen(reinterpret_cast<const char*>(string))); | |
| 896 } | |
| 897 | |
| 898 String String::fromUTF8(const CString& s) | |
| 899 { | |
| 900 return fromUTF8(s.data()); | |
| 901 } | |
| 902 | |
| 903 String String::fromUTF8WithLatin1Fallback(const LChar* string, size_t size) | |
| 904 { | |
| 905 String utf8 = fromUTF8(string, size); | |
| 906 if (!utf8) | |
| 907 return String(string, size); | |
| 908 return utf8; | |
| 909 } | 840 } |
| 910 | 841 |
| 911 // String Operations | 842 // String Operations |
| 912 | 843 |
| 913 static bool isCharacterAllowedInBase(UChar c, int base) | 844 static bool isCharacterAllowedInBase(UChar c, int base) { |
| 914 { | 845 if (c > 0x7F) |
| 915 if (c > 0x7F) | 846 return false; |
| 916 return false; | 847 if (isASCIIDigit(c)) |
| 848 return c - '0' < base; |
| 849 if (isASCIIAlpha(c)) { |
| 850 if (base > 36) |
| 851 base = 36; |
| 852 return (c >= 'a' && c < 'a' + base - 10) || (c >= 'A' && c < 'A' + base - 10
); |
| 853 } |
| 854 return false; |
| 855 } |
| 856 |
| 857 template <typename IntegralType, typename CharType> |
| 858 static inline IntegralType toIntegralType(const CharType* data, size_t length, b
ool* ok, int base) { |
| 859 static const IntegralType integralMax = std::numeric_limits<IntegralType>::max
(); |
| 860 static const bool isSigned = std::numeric_limits<IntegralType>::is_signed; |
| 861 const IntegralType maxMultiplier = integralMax / base; |
| 862 |
| 863 IntegralType value = 0; |
| 864 bool isOk = false; |
| 865 bool isNegative = false; |
| 866 |
| 867 if (!data) |
| 868 goto bye; |
| 869 |
| 870 // skip leading whitespace |
| 871 while (length && isSpaceOrNewline(*data)) { |
| 872 --length; |
| 873 ++data; |
| 874 } |
| 875 |
| 876 if (isSigned && length && *data == '-') { |
| 877 --length; |
| 878 ++data; |
| 879 isNegative = true; |
| 880 } else if (length && *data == '+') { |
| 881 --length; |
| 882 ++data; |
| 883 } |
| 884 |
| 885 if (!length || !isCharacterAllowedInBase(*data, base)) |
| 886 goto bye; |
| 887 |
| 888 while (length && isCharacterAllowedInBase(*data, base)) { |
| 889 --length; |
| 890 IntegralType digitValue; |
| 891 CharType c = *data; |
| 917 if (isASCIIDigit(c)) | 892 if (isASCIIDigit(c)) |
| 918 return c - '0' < base; | 893 digitValue = c - '0'; |
| 919 if (isASCIIAlpha(c)) { | 894 else if (c >= 'a') |
| 920 if (base > 36) | 895 digitValue = c - 'a' + 10; |
| 921 base = 36; | 896 else |
| 922 return (c >= 'a' && c < 'a' + base - 10) | 897 digitValue = c - 'A' + 10; |
| 923 || (c >= 'A' && c < 'A' + base - 10); | 898 |
| 924 } | 899 if (value > maxMultiplier || (value == maxMultiplier && digitValue > (integr
alMax % base) + isNegative)) |
| 925 return false; | 900 goto bye; |
| 926 } | 901 |
| 927 | 902 value = base * value + digitValue; |
| 928 template <typename IntegralType, typename CharType> | 903 ++data; |
| 929 static inline IntegralType toIntegralType(const CharType* data, size_t length, b
ool* ok, int base) | 904 } |
| 930 { | |
| 931 static const IntegralType integralMax = std::numeric_limits<IntegralType>::m
ax(); | |
| 932 static const bool isSigned = std::numeric_limits<IntegralType>::is_signed; | |
| 933 const IntegralType maxMultiplier = integralMax / base; | |
| 934 | |
| 935 IntegralType value = 0; | |
| 936 bool isOk = false; | |
| 937 bool isNegative = false; | |
| 938 | |
| 939 if (!data) | |
| 940 goto bye; | |
| 941 | |
| 942 // skip leading whitespace | |
| 943 while (length && isSpaceOrNewline(*data)) { | |
| 944 --length; | |
| 945 ++data; | |
| 946 } | |
| 947 | |
| 948 if (isSigned && length && *data == '-') { | |
| 949 --length; | |
| 950 ++data; | |
| 951 isNegative = true; | |
| 952 } else if (length && *data == '+') { | |
| 953 --length; | |
| 954 ++data; | |
| 955 } | |
| 956 | |
| 957 if (!length || !isCharacterAllowedInBase(*data, base)) | |
| 958 goto bye; | |
| 959 | |
| 960 while (length && isCharacterAllowedInBase(*data, base)) { | |
| 961 --length; | |
| 962 IntegralType digitValue; | |
| 963 CharType c = *data; | |
| 964 if (isASCIIDigit(c)) | |
| 965 digitValue = c - '0'; | |
| 966 else if (c >= 'a') | |
| 967 digitValue = c - 'a' + 10; | |
| 968 else | |
| 969 digitValue = c - 'A' + 10; | |
| 970 | |
| 971 if (value > maxMultiplier || (value == maxMultiplier && digitValue > (in
tegralMax % base) + isNegative)) | |
| 972 goto bye; | |
| 973 | |
| 974 value = base * value + digitValue; | |
| 975 ++data; | |
| 976 } | |
| 977 | 905 |
| 978 #if COMPILER(MSVC) | 906 #if COMPILER(MSVC) |
| 979 #pragma warning(push, 0) | 907 #pragma warning(push, 0) |
| 980 #pragma warning(disable:4146) | 908 #pragma warning(disable : 4146) |
| 981 #endif | 909 #endif |
| 982 | 910 |
| 983 if (isNegative) | 911 if (isNegative) |
| 984 value = -value; | 912 value = -value; |
| 985 | 913 |
| 986 #if COMPILER(MSVC) | 914 #if COMPILER(MSVC) |
| 987 #pragma warning(pop) | 915 #pragma warning(pop) |
| 988 #endif | 916 #endif |
| 989 | 917 |
| 990 // skip trailing space | 918 // skip trailing space |
| 991 while (length && isSpaceOrNewline(*data)) { | 919 while (length && isSpaceOrNewline(*data)) { |
| 992 --length; | 920 --length; |
| 993 ++data; | 921 ++data; |
| 994 } | 922 } |
| 995 | 923 |
| 996 if (!length) | 924 if (!length) |
| 997 isOk = true; | 925 isOk = true; |
| 998 bye: | 926 bye: |
| 927 if (ok) |
| 928 *ok = isOk; |
| 929 return isOk ? value : 0; |
| 930 } |
| 931 |
| 932 template <typename CharType> |
| 933 static unsigned lengthOfCharactersAsInteger(const CharType* data, size_t length)
{ |
| 934 size_t i = 0; |
| 935 |
| 936 // Allow leading spaces. |
| 937 for (; i != length; ++i) { |
| 938 if (!isSpaceOrNewline(data[i])) |
| 939 break; |
| 940 } |
| 941 |
| 942 // Allow sign. |
| 943 if (i != length && (data[i] == '+' || data[i] == '-')) |
| 944 ++i; |
| 945 |
| 946 // Allow digits. |
| 947 for (; i != length; ++i) { |
| 948 if (!isASCIIDigit(data[i])) |
| 949 break; |
| 950 } |
| 951 |
| 952 return i; |
| 953 } |
| 954 |
| 955 int charactersToIntStrict(const LChar* data, size_t length, bool* ok, int base)
{ |
| 956 return toIntegralType<int, LChar>(data, length, ok, base); |
| 957 } |
| 958 |
| 959 int charactersToIntStrict(const UChar* data, size_t length, bool* ok, int base)
{ |
| 960 return toIntegralType<int, UChar>(data, length, ok, base); |
| 961 } |
| 962 |
| 963 unsigned charactersToUIntStrict(const LChar* data, size_t length, bool* ok, int
base) { |
| 964 return toIntegralType<unsigned, LChar>(data, length, ok, base); |
| 965 } |
| 966 |
| 967 unsigned charactersToUIntStrict(const UChar* data, size_t length, bool* ok, int
base) { |
| 968 return toIntegralType<unsigned, UChar>(data, length, ok, base); |
| 969 } |
| 970 |
| 971 int64_t charactersToInt64Strict(const LChar* data, size_t length, bool* ok, int
base) { |
| 972 return toIntegralType<int64_t, LChar>(data, length, ok, base); |
| 973 } |
| 974 |
| 975 int64_t charactersToInt64Strict(const UChar* data, size_t length, bool* ok, int
base) { |
| 976 return toIntegralType<int64_t, UChar>(data, length, ok, base); |
| 977 } |
| 978 |
| 979 uint64_t charactersToUInt64Strict(const LChar* data, size_t length, bool* ok, in
t base) { |
| 980 return toIntegralType<uint64_t, LChar>(data, length, ok, base); |
| 981 } |
| 982 |
| 983 uint64_t charactersToUInt64Strict(const UChar* data, size_t length, bool* ok, in
t base) { |
| 984 return toIntegralType<uint64_t, UChar>(data, length, ok, base); |
| 985 } |
| 986 |
| 987 int charactersToInt(const LChar* data, size_t length, bool* ok) { |
| 988 return toIntegralType<int, LChar>(data, lengthOfCharactersAsInteger<LChar>(dat
a, length), ok, 10); |
| 989 } |
| 990 |
| 991 int charactersToInt(const UChar* data, size_t length, bool* ok) { |
| 992 return toIntegralType<int, UChar>(data, lengthOfCharactersAsInteger(data, leng
th), ok, 10); |
| 993 } |
| 994 |
| 995 unsigned charactersToUInt(const LChar* data, size_t length, bool* ok) { |
| 996 return toIntegralType<unsigned, LChar>(data, lengthOfCharactersAsInteger<LChar
>(data, length), ok, 10); |
| 997 } |
| 998 |
| 999 unsigned charactersToUInt(const UChar* data, size_t length, bool* ok) { |
| 1000 return toIntegralType<unsigned, UChar>(data, lengthOfCharactersAsInteger<UChar
>(data, length), ok, 10); |
| 1001 } |
| 1002 |
| 1003 int64_t charactersToInt64(const LChar* data, size_t length, bool* ok) { |
| 1004 return toIntegralType<int64_t, LChar>(data, lengthOfCharactersAsInteger<LChar>
(data, length), ok, 10); |
| 1005 } |
| 1006 |
| 1007 int64_t charactersToInt64(const UChar* data, size_t length, bool* ok) { |
| 1008 return toIntegralType<int64_t, UChar>(data, lengthOfCharactersAsInteger<UChar>
(data, length), ok, 10); |
| 1009 } |
| 1010 |
| 1011 uint64_t charactersToUInt64(const LChar* data, size_t length, bool* ok) { |
| 1012 return toIntegralType<uint64_t, LChar>(data, lengthOfCharactersAsInteger<LChar
>(data, length), ok, 10); |
| 1013 } |
| 1014 |
| 1015 uint64_t charactersToUInt64(const UChar* data, size_t length, bool* ok) { |
| 1016 return toIntegralType<uint64_t, UChar>(data, lengthOfCharactersAsInteger<UChar
>(data, length), ok, 10); |
| 1017 } |
| 1018 |
| 1019 enum TrailingJunkPolicy { DisallowTrailingJunk, |
| 1020 AllowTrailingJunk }; |
| 1021 |
| 1022 template <typename CharType, TrailingJunkPolicy policy> |
| 1023 static inline double toDoubleType(const CharType* data, size_t length, bool* ok,
size_t& parsedLength) { |
| 1024 size_t leadingSpacesLength = 0; |
| 1025 while (leadingSpacesLength < length && isASCIISpace(data[leadingSpacesLength])
) |
| 1026 ++leadingSpacesLength; |
| 1027 |
| 1028 double number = parseDouble(data + leadingSpacesLength, length - leadingSpaces
Length, parsedLength); |
| 1029 if (!parsedLength) { |
| 999 if (ok) | 1030 if (ok) |
| 1000 *ok = isOk; | 1031 *ok = false; |
| 1001 return isOk ? value : 0; | 1032 return 0.0; |
| 1002 } | 1033 } |
| 1003 | 1034 |
| 1004 template <typename CharType> | 1035 parsedLength += leadingSpacesLength; |
| 1005 static unsigned lengthOfCharactersAsInteger(const CharType* data, size_t length) | 1036 if (ok) |
| 1006 { | 1037 *ok = policy == AllowTrailingJunk || parsedLength == length; |
| 1007 size_t i = 0; | 1038 return number; |
| 1008 | 1039 } |
| 1009 // Allow leading spaces. | 1040 |
| 1010 for (; i != length; ++i) { | 1041 double charactersToDouble(const LChar* data, size_t length, bool* ok) { |
| 1011 if (!isSpaceOrNewline(data[i])) | 1042 size_t parsedLength; |
| 1012 break; | 1043 return toDoubleType<LChar, DisallowTrailingJunk>(data, length, ok, parsedLengt
h); |
| 1013 } | 1044 } |
| 1014 | 1045 |
| 1015 // Allow sign. | 1046 double charactersToDouble(const UChar* data, size_t length, bool* ok) { |
| 1016 if (i != length && (data[i] == '+' || data[i] == '-')) | 1047 size_t parsedLength; |
| 1017 ++i; | 1048 return toDoubleType<UChar, DisallowTrailingJunk>(data, length, ok, parsedLengt
h); |
| 1018 | 1049 } |
| 1019 // Allow digits. | 1050 |
| 1020 for (; i != length; ++i) { | 1051 float charactersToFloat(const LChar* data, size_t length, bool* ok) { |
| 1021 if (!isASCIIDigit(data[i])) | 1052 // FIXME: This will return ok even when the string fits into a double but |
| 1022 break; | 1053 // not a float. |
| 1023 } | 1054 size_t parsedLength; |
| 1024 | 1055 return static_cast<float>(toDoubleType<LChar, DisallowTrailingJunk>(data, leng
th, ok, parsedLength)); |
| 1025 return i; | 1056 } |
| 1026 } | 1057 |
| 1027 | 1058 float charactersToFloat(const UChar* data, size_t length, bool* ok) { |
| 1028 int charactersToIntStrict(const LChar* data, size_t length, bool* ok, int base) | 1059 // FIXME: This will return ok even when the string fits into a double but |
| 1029 { | 1060 // not a float. |
| 1030 return toIntegralType<int, LChar>(data, length, ok, base); | 1061 size_t parsedLength; |
| 1031 } | 1062 return static_cast<float>(toDoubleType<UChar, DisallowTrailingJunk>(data, leng
th, ok, parsedLength)); |
| 1032 | 1063 } |
| 1033 int charactersToIntStrict(const UChar* data, size_t length, bool* ok, int base) | 1064 |
| 1034 { | 1065 float charactersToFloat(const LChar* data, size_t length, size_t& parsedLength)
{ |
| 1035 return toIntegralType<int, UChar>(data, length, ok, base); | 1066 // FIXME: This will return ok even when the string fits into a double but |
| 1036 } | 1067 // not a float. |
| 1037 | 1068 return static_cast<float>(toDoubleType<LChar, AllowTrailingJunk>(data, length,
0, parsedLength)); |
| 1038 unsigned charactersToUIntStrict(const LChar* data, size_t length, bool* ok, int
base) | 1069 } |
| 1039 { | 1070 |
| 1040 return toIntegralType<unsigned, LChar>(data, length, ok, base); | 1071 float charactersToFloat(const UChar* data, size_t length, size_t& parsedLength)
{ |
| 1041 } | 1072 // FIXME: This will return ok even when the string fits into a double but |
| 1042 | 1073 // not a float. |
| 1043 unsigned charactersToUIntStrict(const UChar* data, size_t length, bool* ok, int
base) | 1074 return static_cast<float>(toDoubleType<UChar, AllowTrailingJunk>(data, length,
0, parsedLength)); |
| 1044 { | 1075 } |
| 1045 return toIntegralType<unsigned, UChar>(data, length, ok, base); | 1076 |
| 1046 } | 1077 const String& emptyString() { |
| 1047 | 1078 DEFINE_STATIC_LOCAL(String, emptyString, (StringImpl::empty())); |
| 1048 int64_t charactersToInt64Strict(const LChar* data, size_t length, bool* ok, int
base) | 1079 return emptyString; |
| 1049 { | 1080 } |
| 1050 return toIntegralType<int64_t, LChar>(data, length, ok, base); | 1081 |
| 1051 } | 1082 const String& emptyString16Bit() { |
| 1052 | 1083 DEFINE_STATIC_LOCAL(String, emptyString, (StringImpl::empty16Bit())); |
| 1053 int64_t charactersToInt64Strict(const UChar* data, size_t length, bool* ok, int
base) | 1084 return emptyString; |
| 1054 { | 1085 } |
| 1055 return toIntegralType<int64_t, UChar>(data, length, ok, base); | 1086 |
| 1056 } | 1087 } // namespace WTF |
| 1057 | |
| 1058 uint64_t charactersToUInt64Strict(const LChar* data, size_t length, bool* ok, in
t base) | |
| 1059 { | |
| 1060 return toIntegralType<uint64_t, LChar>(data, length, ok, base); | |
| 1061 } | |
| 1062 | |
| 1063 uint64_t charactersToUInt64Strict(const UChar* data, size_t length, bool* ok, in
t base) | |
| 1064 { | |
| 1065 return toIntegralType<uint64_t, UChar>(data, length, ok, base); | |
| 1066 } | |
| 1067 | |
| 1068 int charactersToInt(const LChar* data, size_t length, bool* ok) | |
| 1069 { | |
| 1070 return toIntegralType<int, LChar>(data, lengthOfCharactersAsInteger<LChar>(d
ata, length), ok, 10); | |
| 1071 } | |
| 1072 | |
| 1073 int charactersToInt(const UChar* data, size_t length, bool* ok) | |
| 1074 { | |
| 1075 return toIntegralType<int, UChar>(data, lengthOfCharactersAsInteger(data, le
ngth), ok, 10); | |
| 1076 } | |
| 1077 | |
| 1078 unsigned charactersToUInt(const LChar* data, size_t length, bool* ok) | |
| 1079 { | |
| 1080 return toIntegralType<unsigned, LChar>(data, lengthOfCharactersAsInteger<LCh
ar>(data, length), ok, 10); | |
| 1081 } | |
| 1082 | |
| 1083 unsigned charactersToUInt(const UChar* data, size_t length, bool* ok) | |
| 1084 { | |
| 1085 return toIntegralType<unsigned, UChar>(data, lengthOfCharactersAsInteger<UCh
ar>(data, length), ok, 10); | |
| 1086 } | |
| 1087 | |
| 1088 int64_t charactersToInt64(const LChar* data, size_t length, bool* ok) | |
| 1089 { | |
| 1090 return toIntegralType<int64_t, LChar>(data, lengthOfCharactersAsInteger<LCha
r>(data, length), ok, 10); | |
| 1091 } | |
| 1092 | |
| 1093 int64_t charactersToInt64(const UChar* data, size_t length, bool* ok) | |
| 1094 { | |
| 1095 return toIntegralType<int64_t, UChar>(data, lengthOfCharactersAsInteger<UCha
r>(data, length), ok, 10); | |
| 1096 } | |
| 1097 | |
| 1098 uint64_t charactersToUInt64(const LChar* data, size_t length, bool* ok) | |
| 1099 { | |
| 1100 return toIntegralType<uint64_t, LChar>(data, lengthOfCharactersAsInteger<LCh
ar>(data, length), ok, 10); | |
| 1101 } | |
| 1102 | |
| 1103 uint64_t charactersToUInt64(const UChar* data, size_t length, bool* ok) | |
| 1104 { | |
| 1105 return toIntegralType<uint64_t, UChar>(data, lengthOfCharactersAsInteger<UCh
ar>(data, length), ok, 10); | |
| 1106 } | |
| 1107 | |
| 1108 enum TrailingJunkPolicy { DisallowTrailingJunk, AllowTrailingJunk }; | |
| 1109 | |
| 1110 template <typename CharType, TrailingJunkPolicy policy> | |
| 1111 static inline double toDoubleType(const CharType* data, size_t length, bool* ok,
size_t& parsedLength) | |
| 1112 { | |
| 1113 size_t leadingSpacesLength = 0; | |
| 1114 while (leadingSpacesLength < length && isASCIISpace(data[leadingSpacesLength
])) | |
| 1115 ++leadingSpacesLength; | |
| 1116 | |
| 1117 double number = parseDouble(data + leadingSpacesLength, length - leadingSpac
esLength, parsedLength); | |
| 1118 if (!parsedLength) { | |
| 1119 if (ok) | |
| 1120 *ok = false; | |
| 1121 return 0.0; | |
| 1122 } | |
| 1123 | |
| 1124 parsedLength += leadingSpacesLength; | |
| 1125 if (ok) | |
| 1126 *ok = policy == AllowTrailingJunk || parsedLength == length; | |
| 1127 return number; | |
| 1128 } | |
| 1129 | |
| 1130 double charactersToDouble(const LChar* data, size_t length, bool* ok) | |
| 1131 { | |
| 1132 size_t parsedLength; | |
| 1133 return toDoubleType<LChar, DisallowTrailingJunk>(data, length, ok, parsedLen
gth); | |
| 1134 } | |
| 1135 | |
| 1136 double charactersToDouble(const UChar* data, size_t length, bool* ok) | |
| 1137 { | |
| 1138 size_t parsedLength; | |
| 1139 return toDoubleType<UChar, DisallowTrailingJunk>(data, length, ok, parsedLen
gth); | |
| 1140 } | |
| 1141 | |
| 1142 float charactersToFloat(const LChar* data, size_t length, bool* ok) | |
| 1143 { | |
| 1144 // FIXME: This will return ok even when the string fits into a double but | |
| 1145 // not a float. | |
| 1146 size_t parsedLength; | |
| 1147 return static_cast<float>(toDoubleType<LChar, DisallowTrailingJunk>(data, le
ngth, ok, parsedLength)); | |
| 1148 } | |
| 1149 | |
| 1150 float charactersToFloat(const UChar* data, size_t length, bool* ok) | |
| 1151 { | |
| 1152 // FIXME: This will return ok even when the string fits into a double but | |
| 1153 // not a float. | |
| 1154 size_t parsedLength; | |
| 1155 return static_cast<float>(toDoubleType<UChar, DisallowTrailingJunk>(data, le
ngth, ok, parsedLength)); | |
| 1156 } | |
| 1157 | |
| 1158 float charactersToFloat(const LChar* data, size_t length, size_t& parsedLength) | |
| 1159 { | |
| 1160 // FIXME: This will return ok even when the string fits into a double but | |
| 1161 // not a float. | |
| 1162 return static_cast<float>(toDoubleType<LChar, AllowTrailingJunk>(data, lengt
h, 0, parsedLength)); | |
| 1163 } | |
| 1164 | |
| 1165 float charactersToFloat(const UChar* data, size_t length, size_t& parsedLength) | |
| 1166 { | |
| 1167 // FIXME: This will return ok even when the string fits into a double but | |
| 1168 // not a float. | |
| 1169 return static_cast<float>(toDoubleType<UChar, AllowTrailingJunk>(data, lengt
h, 0, parsedLength)); | |
| 1170 } | |
| 1171 | |
| 1172 const String& emptyString() | |
| 1173 { | |
| 1174 DEFINE_STATIC_LOCAL(String, emptyString, (StringImpl::empty())); | |
| 1175 return emptyString; | |
| 1176 } | |
| 1177 | |
| 1178 const String& emptyString16Bit() | |
| 1179 { | |
| 1180 DEFINE_STATIC_LOCAL(String, emptyString, (StringImpl::empty16Bit())); | |
| 1181 return emptyString; | |
| 1182 } | |
| 1183 | |
| 1184 } // namespace WTF | |
| 1185 | 1088 |
| 1186 #ifndef NDEBUG | 1089 #ifndef NDEBUG |
| 1187 // For use in the debugger | 1090 // For use in the debugger |
| 1188 String* string(const char*); | 1091 String* string(const char*); |
| 1189 Vector<char> asciiDebug(StringImpl*); | 1092 Vector<char> asciiDebug(StringImpl*); |
| 1190 Vector<char> asciiDebug(String&); | 1093 Vector<char> asciiDebug(String&); |
| 1191 | 1094 |
| 1192 void String::show() const | 1095 void String::show() const { |
| 1193 { | 1096 dataLogF("%s\n", asciiDebug(impl()).data()); |
| 1194 dataLogF("%s\n", asciiDebug(impl()).data()); | |
| 1195 } | 1097 } |
| 1196 | 1098 |
| 1197 String* string(const char* s) | 1099 String* string(const char* s) { |
| 1198 { | 1100 // leaks memory! |
| 1199 // leaks memory! | 1101 return new String(s); |
| 1200 return new String(s); | |
| 1201 } | 1102 } |
| 1202 | 1103 |
| 1203 Vector<char> asciiDebug(StringImpl* impl) | 1104 Vector<char> asciiDebug(StringImpl* impl) { |
| 1204 { | 1105 if (!impl) |
| 1205 if (!impl) | 1106 return asciiDebug(String("[null]").impl()); |
| 1206 return asciiDebug(String("[null]").impl()); | |
| 1207 | 1107 |
| 1208 Vector<char> buffer; | 1108 Vector<char> buffer; |
| 1209 for (unsigned i = 0; i < impl->length(); ++i) { | 1109 for (unsigned i = 0; i < impl->length(); ++i) { |
| 1210 UChar ch = (*impl)[i]; | 1110 UChar ch = (*impl)[i]; |
| 1211 if (isASCIIPrintable(ch)) { | 1111 if (isASCIIPrintable(ch)) { |
| 1212 if (ch == '\\') | 1112 if (ch == '\\') |
| 1213 buffer.append('\\'); | 1113 buffer.append('\\'); |
| 1214 buffer.append(static_cast<char>(ch)); | 1114 buffer.append(static_cast<char>(ch)); |
| 1215 } else { | 1115 } else { |
| 1216 buffer.append('\\'); | 1116 buffer.append('\\'); |
| 1217 buffer.append('u'); | 1117 buffer.append('u'); |
| 1218 appendUnsignedAsHexFixedSize(ch, buffer, 4); | 1118 appendUnsignedAsHexFixedSize(ch, buffer, 4); |
| 1219 } | |
| 1220 } | 1119 } |
| 1221 buffer.append('\0'); | 1120 } |
| 1222 return buffer; | 1121 buffer.append('\0'); |
| 1122 return buffer; |
| 1223 } | 1123 } |
| 1224 | 1124 |
| 1225 Vector<char> asciiDebug(String& string) | 1125 Vector<char> asciiDebug(String& string) { |
| 1226 { | 1126 return asciiDebug(string.impl()); |
| 1227 return asciiDebug(string.impl()); | |
| 1228 } | 1127 } |
| 1229 | 1128 |
| 1230 #endif | 1129 #endif |
| OLD | NEW |